CEONSS dedicated server test - Saturday Oct 8th, UTC+2 22:00

[McLovin]

For those of us that don't do this often, here is how to connect to an IP:port.





Make sure you don't try to login with your admin credentials. LOL

[trans_lux]

I’ll be there!

[EmanReleipS]

The server is also listed in the public server list as "Imp)o(ster", but is probably far down the list for you peeps. Just make sure you have the check mark for "standard servers only" removed!

Make sure you don't try to login with your admin credentials. LOL

We mutually agreed that Pooty would not send me your admin configuration file with the rest of the server files. That being said, you should be able to log into the webadmin and console using the custom login name and password that I set up for the Imp)o(ster server instead. During the event we will just have to make sure that not all of us are trying to change maps or settings at the same time.

[pooty]

There's admin credentials for that server check pm box.

If you're reading this join the test!

[pooty]

Thanks to everyone who joined. And thanks to Snarf for all his work.

Observations:
No Crashes! Yay!
UTComp ping compensation worked really well IMO. Hardly ever noticed my 120 ping.
Did not observe any missing turrets/lasers.
Didn't noticed any rampant unregs, but hard to tell.. I swear mino gets the most unregs...only played in it a few times.

CPU Usage:
These are all the higher observed values meaning on Bitchslap is was often a bit lower than 7.3%, but think of it as the highwater mark, and pretty much held that steady state, occasionally a 1-3% variance.
BitchSlap - 13 players - 7.3% CPU
RailBeta - 13 players - 7%, 14 players - 9-10%
Minus-KitchenSink - 11 players, 11% CPU, 14 players went to 17%, 15 to 18%, Had 16 Players CPU 20-23%
HSH - 13 players around 13%.

So overall not really hard on the CPU at all with 16 players. Now at 32 we'd probably see 40-50% CPU similar to what we see now.

Last night on our we had 26 players on CPU peaked at 38%, averages between 24-32%. (Not sure what map)

So not sure need dedicated HW, I think fixing the UTComp stuff, plus the slightly improved CPU from NFO might get us there.

[captainsnarf]

net usage was:

Code: Select all

Average
TX: 937Kb
RX: 451Kb
Total: 1.36Mb

Peaks
TX: 1.20Mb
RX: 585Kb
Total: 1.72Mb

Which seems pretty light. Dialup might even handled it

[leon]

Well played. Maybe I'm wrong, but I felt the difference in ping. Tried to kill McLovin, fired 10 missiles at him and he ended up killing me (more than once).

[McLovin]

...
Tried to kill McLovin, fired 10 missiles at him and he ended up killing me (more than once).

It always feels good when I get a leon-kill, which isn't often, LOL

Pooty, is there a UTStats page for this event?

https://omnipotents2.site.nfoservers.co ... ts=matches

[pooty]

Pooty, is there a UTStats page for this event?

No it was just a test, so stats weren't recorded. Too bad, I got some good kills in on leon too. Doesn't happen too often.

[trans_lux]

Yes big thanks to Snarf for making this happen.
I had what felt like significantly more unregistered events.
Many times they were at close range and with weapons that have splash damage.
It felt like there was some consistency with specific players.
They wouldn’t take damage and there was no hit sounds even after what would normally be multiple good hits.
Conversely i was also able to get some good Leon kills so IDK.
If it matters I’m running a 64Bit version at 2560 × 1440 @ 165hz.